Donwshifting to First at Autocrosses
 

Downshifting to First at Autocrosses
This is for all you Autocross freaks out there. I'm one of those autocrossers who
enjoys the simple pleasure of heel and toe downshifting into first gear at autocross
courses. I sometimes find the friendly more experienced people (who are willing to help me
improve) questioning its effectiveness bringing up the common believe that you will lose
about 0.3 seconds everytime you shift. I have even tested it out on a few different
courses with and w/o downshifting. On some courses it does help, on others there's no
difference but I don't think downshifting to first has ever significantly effected my
times negatively. I do admit that sometimes I downshift too often but it's so much fun as
it sounds good and on certain courses, it gets you into a rhythm.

Because of the the questions I get as to whether or not it helps out, I have given it
some thought and decided to do some calculations to put my mind at ease. Here are
a set of simple calculations that shows the effects of downshifting on an autocross course.


The calculations will revolve around a 93 Chevrolet Camaro Z28. The reason I chose this car
was because I had the information handy. Also with it has a killer powerband and I can
assume that it has an almost linear acceleration in the gears I will be doing my
calculations around. So if you are interested in the calculations, go ahead and stretch.
Get yourself a cup of coffee, get ready your calculator, some paper, a pencil (or MathCAD)
Because this is going to be pretty long. If you trust that I didn't make any mistakes,
there will be conclusions at the end and maybe an excel graph in a later post. Please note
that this is not an attempt to make anyone change their driving style but just an attempt
to justify my habit of downshifting. Please point out any mistakes I have made. Alright
here we go.

Here are some specs on the car a 93 Chevrolet Camaro Z28

Acceleration numbers

0-30mph ~ 2.7s
0-40mph ~ 3.6s
0-50mph ~ 4.9s
0-60mph ~ 6.2s

Other Specifications

275hp @ 5200rpm
325lb-ft @ 2400rpm
redline at 5750rpm

245/50/R16 tyres

Gear ratios
1st 3.36
2nd 2.07
3rd 1.35
4th 1.00
5th 0.80
6th 0.62
Final 3.23

Here is a unit conversion that I will use

1mph = 1mile/hour * 1hour/3600s * 5280ft/1mile

1mph = 1.47ft/s

1in = 25.4mm

1ft = 12in duh!

Now to calculate the top speeds of the Z28 for 1st and 2nd gear.

The size of the wheel is 245/50/16 so the diameter of the wheel is

Diameter = (245mm/25.4 * 0.5 * 2)in + 16in
= 25.65in
= 2.14ft

Therefore circumference is

Circumference = Diameter * Pi
= 2.14ft * 3.142
= 6.72ft

Therefore The top speed for 1st gear would be reached at a redline of 5750rpm

Top Speed 1st gear = RedlineRPM/1stgearratio/finaldriveratio/60seconds * Circumference
= 5750rpm/3.36/3.23/60seconds * 6.72ft
= 59.34ft/s
= 40mph (conversion: 1mph = 1.47ft/s)

Top Speed 2nd gear = RedlineRPM/2ndgearratio/finaldriveratio/60seconds * Circumference
= 5750rpm/2.07/3.23/60seconds * 6.72ft
= 96.32ft/s
= 65.52mph (conversion: 1mph = 1.47ft/s)

So now we know that 1st gear goes all the way to 40mph and 2nd gear goes to about 65mph.
Now we will estimate the acceleration of 1st and 2nd gear. From the information extracted
from the information above

1stgear acceleration, a1 = 40mph/3.6seconds (conversion: 1mph = 1.47ft/s)
= (58.8ft/s)/3.6seconds
= 16.33ft/s^2

2ndgear acceleration, a2 = (60mph-40mph)/(6.2seconds-3.6seconds)
= (88.2ft/s - 58.5ft/s)/2.6
= 11.31ft/s^2





Now comes the first question. Is an upshift really add 0.3seconds to your final score?

Professional drivers are known to shift as quickly as 0.2 seconds (with racing transmissions
as low as 0.14s!! I think) while really good drivers can shift at about 0.23 seconds I'm
somewhat of an incompetent shifter so lets say I shift at 0.3seconds.

Lets say that with the Camaro we shift at 40mph and while shifting you lose some speed
so let us assume an average speed of 39mph while shifting.

Shifting speed = 39mph (conversion: 1mph = 1.47ft/s)
= 57.33ft/s

Lets say I take 0.3 seconds to complete a shift so therefore the distance I will travel
while making the shift is

ShiftDistance = 57.33ft/s * 0.3s
= 17.2ft

WOW! You travel 17ft while shifting from 1st to 2nd!

Now, Assume that you were already in 2nd gear when you were at 40mph. That means you have
0.3seconds of acceleration advantage over the gear shifting person. The speed gained if you
accelerated in 2nd gear at 11.31ft/s is


Speed gained = 2nd gear acceleration * 0.3seconds
= 11.31ft/s * 0.3seconds
= 3.39ft/s (conversion: 1mph = 1.47ft/s)
= 2.31mph

Final speed after 0.3 seconds

Finalspeed = 40mph + 2.31mph
= 42.31mph (conversion: 1mph = 1.47ft/s)
= 62.19ft/s

Therefore the distance you will travel in the 0.3seconds if you were already in 2nd gear
and accelerating is

Noshiftdistance = (Finalspeed + initial speed)/2 * 0.3s
= (62.19ft/s + 58.8ft/s)/2 * 0.3s
= 18.15ft

So the distance an upshift will cost you is

Lost distance = Noshiftdistance - Shiftdistance
= 18.15ft - 17.2ft
= 0.95ft

You will lose almost a feet to the person who is did not downshift but who is already in 2nd.
But is this 1 feet really worth 0.3seconds? Assuming that you are travelling about
40mph(duh!) the time lost due to shifting would be

Time lost = lost distance/speed
= 0.95ft/40mph (conversion: 1mph = 1.47ft/s)
= 0.95ft/(58.8ft/s)
= 0.016seconds

So the time lost would only be 0.016seconds.. There just as I suspected... the time lost
due to an upshift has been overestimated. Remember the common believe that you lose
0.3 seconds with every upshift? Well.. that would be true if the car instantaneously
stopped moving for 0.3 seconds and instantaneously starts moving again at 40mph! But the
car is constantly in motion thus cutting your losses.

Okay... now we know that you don't actually lose 0.3 seconds everytime you shift... so what?
that proves that downshifting sucks right? Well... there is a reason you downshift... you
downshift so that you can make better use of the powerband. But in the calculations that
I am doing I'm assuming linear acceleration(So we are being conservative) in 1st and second
so, over here you are doing it to take advantage of the gear ratios.

Let's consider a situation where you take a tight corner and exit it at 25mph. Lets figure
out the time it takes for the Z28 to reach 50mph if it were in second gear all the way. This
situation would be typical of an autocross course. Here we assume that the acceleration is
the same as calculated above(in reality it would be slower since at 25mph the car
would not be in it's optimum powerband)

Total time = 50mph - 25mph/acceleration
= (73.5ft/s - 36.75ft/s)/11.31ft/s^2 (conversion: 1mph = 1.47ft/s)
= 3.25seconds

Distance travelled w/o shifting = (50mph + 25mph)/2 * Total time
= (73.5ft/s + 36.75ft/s)/2 * 3.25 (conversion: 1mph = 1.47ft/s)
= 179.16ft

Alright now for comparison I will calculate the distance the car will travel in 3.25seconds
and the time it takes to travel 179.16ft the driver had downshifted into first.

1st gear acceleration was calculated above. So the time to accelerate from 25mph to 40mph, T1 is

T1 = (40mph - 25mph)/ 16.33 ft/s^2
= (58.8ft/s - 36.75ft/s)/16.33ft/s^2
= 1.35s

Now calculate the distance it will travel accelerating from 25mph to the shifting point
of 40mph. First you calculate the average speed.

Avgspeed1 = (25mph + 40mph)/2
= 32.5mph (conversion: 1mph = 1.47ft/s)
= 47.78ft/s


Therefore

Distance1 = Avgspeed1 * T1
= 47.78ft/s * 1.35s
= 64.5ft

The time it takes to shift was determined earlier to be 0.3s (alright Granny) And
assuming an average speed of 39mph in between shifts, the distance travelled in between
shifts is 17.2ft


Now.. first we want to find out the distance the car travels in 3.25seconds
under maximum accelerating from 25mph if the driver had downshifted into first.
So far from the calculations above the driver has taken up 1.35secs accelerating from 25mph
to 40mph and 0.3secs granny shifting. So the amount of time left to accelerate in
2nd gear from 39mph is

T2 = 3.25secs - 1.35secs - 0.3secs
= 1.6secs

The final speed of the car after accelerating from 39mph (57.33ft/s) is

Final speed = initial speed + 2ndgearAcceleration*time
= 57.33ft/s + (11.31ft/s^2*1.6s)
= 75.43ft/s
= 51.31mph

And the distance travelled in that final 1.6secs is

Distance2 = (75.43ft/s + 57.33ft/s)/2 * 1.6secs
= 106.21ft

Totaldistance = Distance2 + Distancegrannyshift + Distance1
= 106.21ft + 17.2ft + 64.5ft
= 187.91


So, in 3.25 seconds starting at a speed of 25mph, the car will reach a speed of 51.31mph and
would have travelled a distance of 187.91. In comparison,the other car which started
accelerating in 2nd gear from 25mph only reached 50mph and travelled 179.16ft. So by
downshifting, the exit speed and distance travelled is increased.

On another note it would take the car about 3.12seconds to reach the distance of
179.16ft if the driver had downshifted. It would take the car 3.25secs to reach that
distance had the driver not downshifted.

Conclusions

- You will not lose 0.3seconds by upshifting. You are more likely to lose somewhere around
0.02seconds
- Accelerating a Z28 from 25mph to 50mph, the person that downshifts to first will be faster

Person who doesn't downshifts
- 25mph to 50mph in 3.25seconds and travel 179.16ft

Person who downshifts taking into consideration the time lost by upshifting
- will travel from 25mph to 51.31mph in 3.25seconds
- Will only take 3.12seconds to reach 179.16ft

Therefore now we know that downshifting at the right times does help!

Other Considerations

- By downshifting you might lose the sensitivity you have when you don't
downshift and concentrate on driving.
- These are the gains you can expect to get IF YOU USE HEEL & TOE downshifting.
- No one is perfect when using heel & toe.. expect to lose some time heel and toeing (I think it's negligible)
- These calculations revolved around a car with a nice powerband. Cars with peaky engines
stand a chance to benefit more from downshifting.
- You can't utilize full acceleration when you there's a turn immediately after downshifting.
- YOU MUST BE SMOOOOOOOOOOOOOOOOOTH

Nice write up.

I don't know if you visit www.sccaforums.com at all but you might want to post this over there also. You will either be flamed or praised. There are some very technical guys over there that have actually probably already done this.

I think the key for most autocrosses is to be smooth as you state. For someone without 100s of hours practicing a heal-toe it is hard be smooth. You either jab the brakes too hard or not enough then have to give it a little more brake to get the speed down. That is where you lose the time. Upshifts are easy good downshifts are hard.

So with that said I still think it is good advice to just stick it in second and go. Unless, you can heal-toe like a pro.

The real question for me is does it make sense to shift to third in a fast section then back to second for turns. Many times I am running at 8-8.5k rpms in second but for such a short time it really just doesn't make sense to shift.

OKay cool.. I will post it up there. I don't mind being flamed. I'm sure there are things in there I ddn't consider. Well... about heel and toe.. I think if you keep at it and practice practice practice one should be able to get it to where any losses associated with it will be negligible. I don't think you need to be as good as a pro to reap the benefits.
Even the best drivers acknowledge that they don't heel and toe good enough... and they back it up with DAQ results.

The problem for me is that I can't practice heal-toeing daily in my "race car". My daily driver is a Saturn (always 35+ MPG, and no expenses other than brakes and oil so far) with bad brakes and a worse clutch so heal-toeing is easy. I am good at it. When it's time to do it in the RX-7 with stainless brake lines, carbotech pads and a decent clutch it is a completly different feel. I can usually do it very well after a while on a track but to do it at an autocross is a little different.

As far as DAQ is concerned I was going to mention that it would be great to see how what you described actually shows up in reality. Once I get my system back up and running it may be time to do some experimenting. My guess is that it will be a very even wash. So then the question is why make it harder on yourself than you need to.

Still great to see some theoretical data on the issue though.

The key to being fast on an autocross course, is being smooth. The fast guys don't look that fast because they don't generally upset the car.

Your example of a Z28 is a poor one. A car with a torquey V-8 is the one that is least likely to need to downshift. That car will have plenty of power at lower rpm's in second gear. It will not need to downshift unless there is a tight hairpin turn on the course.

I have been on some tough, ugly courses with my FD (turbo lag at low RPM!) and I have downshifted to first approximately twice in over 4 years. Why? It doesn't do any good to spend a moment downshifting to first, zing the engine to redline (which happens practically instantly in first gear when the car is already moving) while trying to modulate the throttle so the rearend doesn't walk all over the place and then still have to make another shift back into second anyway. There's no sense in it unless you have a long first gear, otherwise you're just creating more work to do in the car. It's no different than holding a gear rather than upshifting. Why upshift and then have to shift back down 20 feet later? Just hold the gear you've got; you'll come out ahead.

The time you spend actually making the shift doesn't equate to time lost on the course as you've already stated. The few moments it takes to execute the shift do not add into the laptime because the car doesn't come to a halt when you shift, it's still moving. OTOH time spent shifting and aggressively accelerating the car after shifts upsets the car. The far majority of the time I find it quicker to not bother downshifting and just wait for the engine to come on song.

Again this all depends on the gear spacing of the car. The Z06's I run against frequently run the majority of small courses in first gear whereas I am in second the whole time. When they're at the top of second I'm well into third. Different horses for different courses.

Time lost do a shift is not due to the act of moving the clutch and shifter, it's due to the extra workload and the upset of the car. Also don't forget what happens if you should miss the shift; then you've really blown it.

Why is the example of a Z28 a poor one? This is a car with a healthy powerband and the calculations demonstrate that even with a car with a healthy powerband can benefit from a downshift. In fact these calculations has given the this z28 and added advantage. The Z28 will not accelerate as fast as in the calculations from 25mph to 40mph in second gear because it's not in it's optimum powerband.

The car has plenty of torque at lower rpms... that is not to be mistaken for HP. It makes 325ft-lbs @2400rpms punching that in to your calculator you will get

HP = 325 * 2400 / 5252
=148.5hp

not much power considering a heavy car.

Damn... I need to get a race car :(


Well... I don't remember where, but I've seen DAQ's demonstrating that the brake peddle pressure is relieved ever so slightly when professional drivers perform the heel and toe manouever. Because the car is still moving time lost will be minimal. However, I think time lost will only be in the thousands.. even if you lost a few hundreths, from the calculations, you will still make up the time.



hehe yea.. nice to see things in a different perspective once in a while

I think that you are missing something. The SCCA forums guys mentioned it. If you are traction limited then going to a lower gear will just cause more wheel spin and more lost time. The z-28 probably falls into the traction limited category.

Well... if you are wrestling the car back into control and if there is a turn immediately after the downshift it probably wont be beneficial.. I think this was lightly touched in the "other items" section, but if you can go WOT and then shift to 2nd.. I don't see why the extra work isn't beneficial. Remember what feels fast isn't always necessarily fast.. it might feel slow stepping your rear slightly sideways for a few fractions of a second but you might make it up with the added acceleration.. anyway if you do things right and be smooth you shouldn't be stepping the rear end sideways anyway



Well because the Theoretical Calculations indicate a possible reduction in time on course. Of course, there are other outside factors that come into play and this might not always be true... but I think this is true for most cases


Well first gear in Rx7s should top out at about 36mph... I think due to differences in powerbands.. the rx7 should benefit from a downshift if excecuted correctly at the right time

Well I assume that those factors are negligible.. merely a few thousands... even if they cause a few hundreths or even a tenth it would still be beneficial to downshift. Well. this is autocross.. if you miss a shift and ddn't blow your transmission... you will have another 2 or 3 more runs :D

Oh yea.. I forgot to mention a few other items.. but one of them is that lighter cars might not benefit as much from this

Why?

I don't have that many auto-x under my belt (7-8 or so), but I can tell you that it's almost always better to leave it in 2nd gear. I've ran two courses with 180 deg hairpins on slippery pavement (and I have street tires) where a shift to 1st definitely was faster. However, that kind of situation is not typical in my experience. A lot of sections of the course are in fairly low rpm and 1st may be a hair faster perhaps, but going to 1st would also make the car harder to drive and may interfere with steering inputs.

As others have pointed out, your example assumes perfect straightline acceleration -- where in reality, you are always accelerating while coming out of the corner. There are very few corners where I actually NEED more power than WOT in 2nd gear can give me.

Also, for those running race tires, 1st gear becomes even less necessary because of higher cornering speeds (albeit not much higher in a hairpin).

Slacker7

A lot of things work well in theory. However, real world application of many theories proves them to be wrong. The problem with shifting is that it upsets the balance of the car. On an autocross course, you are almost always turning. When you are in a turn, at the limit of traction, shifting will upset the balance of the car, and push you away from the line you are trying to drive. The only time that you can shift without upsetting the car is when you are going pretty straight.

My point about the Z28 is that it is a pretty powerful car. In an autocross, you are almost always using your available traction for cornering. This doesn't leave much available traction for acceleration. A Z28 with lots of power at low rpms, will most likely spin the tires accelerating out of most corners in 2nd gear. If that is the case, there is absolutely no reason to waste time shifting, and unbalance the car at the same time.

You may think that downshifting is faster, and your "theoretical world" calculations may have convinced you. Experience, and the clock, will tell you differently :)

Err.. It varies from car to car.. I was thinking that in lighter cars you normally have a lower capacity engine and with optimized gearing, a lower difference in acceleration rates between gears also it's easier to accelerate closer to full acceleration when you are not in the powerband because the car is lighter. I have some information on an nissan sentra SER handy I might crunch more no's and see what turns out.

I read this several times. I still don't know what you are trying to say. I think you need to spend some time at the track. Also, there are a number of books on autocrossing that will give you some good solid information. I suggest that you do some reading there, before coming up with too many additional theories.

I tend to agree with Damon, and will add:

1) Modulating the throttle in first gear is EXTREMELY hard. I have 350 lf-ft or TQ available at 3k RPM. If you aren't going in a perfectly straight line, it WILL spin 'em. A LOT. By dropping it to second, it allows me a larger margin of error before things get out of hand.
2) Additionally, there are very few situations where I can actually put down more power than WOT in second gear. As mentioned above, there is an obscene amount of power available with that little pedal on the right. In most corners, I'm trying to modulate the throttle to keep the nose tucked (prevent understeer) without cause throttle-induced oversteer. Unless there's a REALLY long straight after the corner, there's no way I'll be able to make use of the gearing advantage of first gear. Second gear, however, seems to work really well. I can go WOT earlier without oversteering.
3) The whole "balance" issue.

And the most important thing:

4) the added complexity. With another shift, you introduce the potential to make a huge mistake. You might miss the rev match which will screw up the balance. You have to perfectly balance the brakes while you do it. You have to not miss the shift. So many "potential" problems that just make it more of a hassle than the 0.1s "potential" gain.

Glad to hear that you agree with ................ Damon :)

Well alot of good theories get thrown out because of the "real world application" excuse. I already have addressed the turning after a downshift problem in the "other considerations" section where I said you might not be able to utilize full acceleration if you are turning and accelerating.... maybe I should have been more explicit. That's why I said in the post if you read it.. It says that downshifting to first gear if done correctly and at the right time will help improve your times. I also explicitly said you need to be smooth.


A lot of the time when you are comming out of a turn in an autocross as 25mph you are not fully utilizing the available traction. This is why you have to note that the throttle is not an "on/of" switch.. it can be controlled if you gradually feed in the throttle. In fact I've noticed normally when comming out of a 25mph turn I have near to ample traction so I gradually feen the throttle. Maybe I'm taking the corner to slow?

If you read my post fully you'd have noticed that experience and the clock has in fact indicated that downshifting actually helps thank you...

Don't we all :D

Been there done that....... and please I think you might understand that statement above better if you went through the 2 main posts and calculations again.

I agree with you, thats why whenever I get into first gear I make sure I can wrap it out fully w/o letting off till the shift (an unsettling the car as little as possible). 2nd gear I also agree will give you a larger margin of error.. but then if you can master the downshift and keep the car settled I think it's worth while taking the risk... if you call it a risk to downshift.


I'm willing to try to learn how to shave an extra 0.1s consistently :)

I read your calculations, and posts. What I (and everyone else) is saying is that your theory doesn't work in the real world. DamonB competes at a national level in his FD. I have been autocrossing, racing in time trials, and have competed in many Pro Solo regional events, as well as SCCA regional finals for more than 20 years.

The quality of your writing assures me that you are not an idiot, so I won't treat you like one. I suggest that you spend some time talking to the experienced guys where you race. See what they tell you .............. since you don't believe me ........... or DamonB :)

Well... sorry if seem to have came off on a bad note.. I ddn't mean to, I just wanted a friendly heated debate. The reason I came up with this is because I have had alot experienced guys telling me what you and DamonB have been telling me. That's including John Thomas. But with my experience and observation it seems that downshifting does help sometimes and others doesn't significantly effect my times. But then again this is with my TurboII which has a peaky powerband(not a good excuse since you guys are prettymuch on the same boat). I just wanted a justification to why I do what I do... That's just me.. being stubborn and defiant in my youth

Next time you go to a race, try it both ways. I think you will eventually find that not shifting is faster. In addition to being faster, it is far easier on your car to stay in one gear. The constant jamming of gears back and forth, in a race, will take its toll on your drive train.

well.. I've already tried it both ways... runs which I stayed in 2nd gear felt alot quicker but it turns out that a single run where I "accidently" downshifted into first for a single corner or a few corners turned out to be slightly faster. I agree staying in 2nd gear all the way is the easiest and there are some courses that I couldn't find a spot to downshift(Good flowing ones). If you are jamming gears back and forth then you are not doing things right.. Isn't one of the purposes of the heel and toe manouever to relieve the transmission of unecessary stresses?